Virtual power plant full-process credible aggregation method and system based on hierarchical trust chain

Abstract

The invention belongs to the technical field of novel electric power system operation control and trust management, and particularly discloses a virtual power plant full-process trusted aggregation method and system based on a hierarchical trust chain. Differentiated hierarchical trust chain models of a data acquisition link, a scheduling control link and a market transaction link are constructed respectively; through real-time credibility evaluation and adaptive weight adjustment, credibility calculation of multi-agent collaborative decision is realized. According to the invention, a trust network formed by a plurality of efficient, safe and transparent virtual power plant trust chains is constructed, solid technical support is provided for fair competition, intelligent scheduling and reliable operation of a power market, and efficient interaction and stable operation of a power system are ensured.

Description

Technical Field

[0001] This invention relates to the field of novel power system operation control and trust management technology, and in particular to a method and system for full-process trusted aggregation of virtual power plants based on hierarchical trust chains. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] Virtual power plants (VPS), as a crucial technological means for integrating distributed energy resources, play a key role in building a new power system dominated by new energy sources. However, with the continuous expansion of distributed energy integration, VPS involve multiple business processes and stakeholders, including data acquisition, dispatch control, and market transactions, resulting in complex information exchange and trust transfer relationships between these processes. The diversity of stakeholders, their wide geographical distribution, and the openness of communication networks lead to serious trust deficiencies, particularly trust gaps during heterogeneous resource integration, which have become significant factors hindering the safe and stable operation of VPS. The lack of a unified trust management mechanism makes it difficult to establish effective trust relationships among stakeholders, resulting in issues such as difficulty in ensuring data authenticity, poor execution of dispatch instructions, and the risk of fraud in market transactions.

[0004] Traditional trust management methods for virtual power plants typically rely on single-point verification or simple identity authentication, paying little attention to the continuity and consistency of trust transfer across different stages, resulting in an incomplete trust chain. Although existing technologies have proposed some trust verification methods based on blockchain and the Internet of Things, these methods often fail to fully consider the differentiated needs of different business stages within a virtual power plant, and a unified trust verification mechanism is difficult to adapt to complex business scenarios. Furthermore, single verification methods often cannot cover the full-stack trust needs from hardware devices to application services, resulting in trust gaps. Moreover, most existing trust models are based on static verification, making it difficult to adapt to the differentiated needs of different business stages within a virtual power plant, lacking targeted trust transfer mechanism design, and failing to achieve an effective combination of real-time trust assessment and long-term trust management. In addition, existing methods lack real-time trust assessment and adaptive adjustment capabilities when facing abnormal situations such as network attacks and equipment failures. Trust management in each business stage is relatively independent, lacking a unified trust transfer and coordination mechanism. Summary of the Invention

[0005] To address the aforementioned issues, this invention proposes a virtual power plant full-process trusted aggregation method and system based on a hierarchical trust chain. This method can construct differentiated trust transfer mechanisms for the characteristic requirements of different business links, establish a complete trust verification system from hardware to applications, realize dynamic trust transfer across links and trusted assurance of multi-subject collaborative decision-making, and effectively solve the trust gap problem in heterogeneous resource integration.

[0006] In some implementations, the following technical solutions are adopted: A method for end-to-end trusted aggregation of a virtual power plant based on a hierarchical trust chain includes: Based on the characteristic requirements and trust transmission patterns of different business segments of a virtual power plant, differentiated hierarchical trust chain models are constructed for the data acquisition segment, the dispatch control segment, and the market transaction segment, respectively. In the data acquisition phase, raw data from terminal devices is collected through edge gateways, and device identity authentication and data integrity verification are combined to ensure the trustworthiness of the underlying devices. The scheduling and control stage receives data streams collected from the edge gateway, predicts photovoltaic and load data, and generates the optimal scheduling scheme. The optimization objective is to minimize the interaction cost with the power grid to maximize the revenue of the virtual power plant. The power balance constraint compliance and energy balance are calculated respectively, and then the reliability of the scheduling scheme is calculated to ensure the reliability of the intermediate layer service. In the market transaction process, buy and sell orders are generated according to the scheduling plan, and a hash value is added to each transaction. The blockchain smart contract verification mechanism is used to realize the reliable execution of the top-level business. The credibility calculation of multi-agent collaborative decision-making is achieved through real-time trust assessment and adaptive weight adjustment.

[0007] In other embodiments, the following technical solutions are adopted: A virtual power plant end-to-end trusted aggregation system based on a hierarchical trust chain includes: The model building module is configured to: build differentiated hierarchical trust chain models for the data acquisition link, the dispatch control link, and the market transaction link based on the characteristic requirements and trust transmission rules of different business links of the virtual power plant; The data acquisition and trust verification module is configured to: collect raw data from terminal devices through the edge gateway during the data acquisition process, and combine device identity authentication and data integrity verification to ensure the trustworthiness of the underlying devices; The scheduling control trust verification module is configured to: receive data streams collected from the edge gateway in the scheduling control process, predict photovoltaic and load, and generate the optimal scheduling scheme; with the goal of minimizing the interaction cost with the power grid, calculate the power balance constraint compliance degree and energy balance degree respectively, and then calculate the trustworthiness of the scheduling scheme to ensure the trustworthiness of the intermediate layer service; The market transaction trust verification module is configured to generate buy and sell orders according to the scheduling scheme during the market transaction process, add a hash value to each transaction, and use a blockchain smart contract verification mechanism to achieve trustworthy execution of the top-level business. The collaborative decision-making credibility calculation module is configured to calculate the credibility of multi-agent collaborative decision-making through real-time trust assessment and adaptive weight adjustment.

[0008] In other embodiments, the following technical solutions are adopted: A terminal device includes a processor and a memory, the processor being used to implement instructions; the memory being used to store multiple instructions, the instructions being adapted to be loaded and executed by the processor to perform the aforementioned trusted aggregation method for the entire process of a virtual power plant based on a hierarchical trust chain.

[0009] In other embodiments, the following technical solutions are adopted: A computer-readable storage medium storing a plurality of instructions adapted for loading and execution by a processor of a terminal device of the aforementioned trusted aggregation method for the entire process of a virtual power plant based on a hierarchical trust chain.

[0010] Compared with the prior art, the beneficial effects of the present invention are: (1) This invention designs a trusted verification system for the aggregated operation of virtual power plants. The virtual power plants not only enable efficient monitoring and resource optimization of the power grid operation, but also ensure the security, transparency, and immutability of information flow through blockchain technology, thereby solving the trust issues between virtual power plants and the power grid dispatching system, as well as in the electricity market. Ultimately, a trust network composed of multiple efficient, secure, and transparent virtual power plant trust chains is constructed, providing solid technical support for fair competition, intelligent dispatching, and reliable operation in the electricity market, ensuring efficient interaction and stable operation of the power system.

[0011] (2) Based on the characteristic requirements of different business links of the virtual power plant, the present invention first constructs a differentiated hierarchical trust chain model, and adopts tree, star and distributed P2P structures to achieve accurate matching between the trust transmission mechanism and the business process; then, it designs a three-level trusted verification system that integrates IoT device authentication, cloud computing dynamic optimization and blockchain smart contracts to establish a complete trust chain from hardware trust root to application layer; finally, through the cross-link dynamic trust transmission mechanism, combined with real-time trust degree assessment and adaptive weight adjustment, it realizes the trusted guarantee of multi-subject collaborative decision-making and constructs a full-process trusted aggregation optimization model for the virtual power plant.

[0012] (3) The method of the present invention can effectively solve the trust gap problem in the integration of heterogeneous resources, significantly improve the trust level and economic benefits of virtual power plants in network attack scenarios, and realize the secure and reliable aggregation of distributed energy resources through a dynamic trust management mechanism. In addition, the method of the present invention has good adaptability and scalability, and can adapt to the operation scenarios and security requirements of virtual power plants of different sizes, providing important technical support for building a new type of power system with new energy as the main body.

[0013] Other features and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the full-process trusted aggregation method for virtual power plants based on a hierarchical trust chain in an embodiment of the present invention; Figure 2 This is a schematic diagram illustrating the generation of a scheduling plan for the next 24 hours in an embodiment of the present invention; Figure 3 This is a schematic diagram of the IEEE 30-node network attack topology in an embodiment of the present invention; Figure 4 This is a schematic diagram comparing the trust enhancement effects in embodiments of the present invention; Figure 5 This is a schematic diagram illustrating how the application of a full-process trust chain in a virtual power plant can optimize and solve problems in this embodiment of the invention. Detailed Implementation

[0015] It should be noted that the following detailed description is illustrative and intended to provide further explanation of the invention. Unless otherwise specified, all technical and scientific terms used in this invention have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0016] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0017] Example 1 In one or more embodiments, a method for end-to-end trusted aggregation of a virtual power plant based on a hierarchical trust chain is disclosed, specifically including the following process: S101: Based on the characteristic requirements and trust transmission rules of different business links of virtual power plants, construct differentiated hierarchical trust chain models for data acquisition, dispatch control and market transaction links respectively; S102: In the data acquisition stage, the edge gateway collects the raw data of the terminal device, and combines device identity authentication and data integrity verification to ensure the trustworthiness of the underlying device; S103: In the scheduling and control stage, the data stream collected from the edge gateway is received, the photovoltaic and load are predicted, and the optimal scheduling scheme is generated. The optimization objective is to minimize the interaction cost with the power grid so as to maximize the revenue of the virtual power plant. The power balance constraint compliance degree and energy balance degree are calculated respectively, and then the scheduling scheme credibility is calculated to ensure the credibility of the intermediate layer service. S104: In the market transaction process, buy and sell orders are generated according to the scheduling plan, and a hash value is added to each transaction. The blockchain smart contract verification mechanism is used to realize the reliable execution of the top-level business. S105: Realize the credibility calculation of multi-agent collaborative decision-making through real-time trust assessment and adaptive weight adjustment.

[0018] In this embodiment, the full-process trusted aggregation method for a virtual power plant based on a hierarchical trust chain mainly includes three core steps: constructing a differentiated hierarchical trust chain, designing a three-level trusted verification system, and establishing a cross-stage dynamic trust transfer mechanism. First, based on the characteristic requirements and trust transfer patterns of different business stages in the virtual power plant, a differentiated hierarchical trust chain model is constructed to accurately match the business processes and trust requirements of each stage. Combined with... Figure 1 To address the massive number of devices accessing the data collection stage, a tree-like hierarchical cascaded structure is adopted to achieve trusted access for IoT devices; to address the centralized management needs of the scheduling and control stage, a star-shaped structure is adopted with the scheduling center at its core to establish a trust management mechanism; and to address the multi-party negotiation characteristics of the market transaction stage, a distributed P2P structure is adopted to achieve decentralized trust verification.

[0019] In this embodiment, a three-tiered trusted verification system is designed to establish a complete trust chain from the hardware root of trust to the application layer. Specifically, a hardware root of trust verification mechanism is built at the device layer, combining device identity authentication and data integrity verification to ensure the trustworthiness of underlying devices. At the platform layer, a cloud computing dynamic optimization verification mechanism is established, employing real-time performance monitoring and anomaly detection algorithms to ensure the trustworthiness of intermediate layer services. At the application layer, a blockchain smart contract verification mechanism is built, utilizing distributed ledger technology to achieve trusted execution of top-level business processes. Finally, a cross-stage dynamic trust transfer mechanism is constructed, achieving trusted assurance for multi-party collaborative decision-making through real-time trust assessment and adaptive weight adjustment.

[0020] As a specific implementation method, the process of constructing a differentiated hierarchical trust chain model is as follows: Starting from the characteristic requirements and trust transmission patterns of different business links in a virtual power plant, this paper constructs a differentiated hierarchical trust chain model for a virtual power plant by analyzing the business characteristics of the three links of data collection, scheduling control, and market transactions, as well as trust transmission structures such as tree, star, and distributed P2P.

[0021] Combination Figure 1 The specific process includes: (1) Overall architecture design of the trust transfer mechanism: This paper analyzes the trust issues in three parts of virtual power plants: data acquisition, dispatch instructions, and market transactions, and constructs a trust transmission mechanism.

[0022] Root of Trust: A Trusted Platform Module (TPM) is established as the hardware root of trust in the central control area of ​​the virtual power plant. As the core hub of the virtual power plant, the central control area not only undertakes the unified scheduling of distributed resources but also serves as the central node for security policy formulation and execution. Establishing a root of trust in this area ensures in-depth defense of system security from the architectural design level, while improving management efficiency and standardization, thus building a reliable security foundation for the entire virtual power plant system.

[0023] Transmission Mode: At the implementation level, the data acquisition chain adopts a tree-like structure, which fully leverages the scalability of tree-like trust chains. By dynamically adjusting the acquisition range and combining it with a hierarchical management mechanism, efficient data aggregation and graded filtering are achieved. This structure supports proximity-based acquisition and data preprocessing strategies, significantly reducing network congestion risks by optimizing transmission paths. The deployment of multi-level data acquisition nodes further enhances the system's real-time performance and reliability, effectively meeting the core technical requirements of the data acquisition process.

[0024] The scheduling and control chain is implemented using a star topology, which allows the central node of the star trust chain to establish direct connections with each execution node. The shortest communication path and lowest transmission latency ensure rapid transmission of control commands. Furthermore, its centralized management mode provides convenient fault diagnosis capabilities and a rapid response mechanism, significantly improving the reliability of the control system and fully meeting the stringent requirements of scheduling and control for low latency, scalability, and flexibility.

[0025] The market transaction chain adopts a distributed P2P structure, a typical implementation of the hybrid dynamic model, which fully reflects the pursuit of transaction transparency. The decentralized nature of the P2P structure effectively eliminates the risk of single points of failure, improves system availability, and strengthens transaction fairness. This structure possesses significant advantages such as multi-point data backup and high fault tolerance, ensuring stable system operation. More importantly, its mechanism of equal participation and information sharing among nodes provides an ideal platform for the implementation of a consensus mechanism, fundamentally guaranteeing transaction transparency and reliability, and aligning with the core technological requirements of modern trading platforms.

[0026] (2) Design of a trusted aggregation operation system for the entire virtual power plant chain: Based on the aforementioned trust chain theoretical framework, a complete trusted verification scheme for the aggregated operation of virtual power plants was designed. This scheme combines edge computing, distributed storage, and blockchain technology to achieve efficient access and trusted management of large-scale distributed energy resources, such as... Figure 1 As shown.

[0027] (2-1) Data Acquisition Trust Verification Subsystem: The data acquisition and trusted verification subsystem is primarily responsible for device access and edge computing, including terminal devices such as smart meters, energy storage PCS, photovoltaic inverters, and load controllers. These devices connect to the system via edge gateways, enabling the system to preprocess and analyze raw data in real time, and to achieve local data storage and management. This improves the system's real-time performance and reliability.

[0028] (2-2) Scheduling and Control Trusted Verification Subsystem: As the data processing hub of the system, a data acquisition service is used to uniformly receive data streams from the edge layer and store them in the InfluxDB distributed time-series database cluster. Photovoltaic power and load forecasts are performed to provide decision-making basis for the scheduling optimization module. The scheduling optimization module comprehensively considers multiple factors such as load forecast results, grid constraints, and economics to generate the optimal scheduling scheme, realizing the intelligent operation of the virtual power plant.

[0029] (2-3) Market Transaction Trust Verification Subsystem: A trustworthy management platform is built using blockchain technology, ensuring the immutability and end-to-end traceability of transaction data through distributed ledger technology. The system integrates energy order management, an automated matching engine, and a blockchain confirmation mechanism to achieve transparent trading and settlement of energy assets. An energy blockchain adapter connects traditional energy dispatching with the blockchain network, ensuring consistency and security in the transaction process. This provides a highly efficient, fair, and reliable trading environment for participants in the distributed energy market, supporting efficient decision-making and system maintenance.

[0030] In this embodiment, the trust chain has the following important characteristics: 1) Verifiability: Any entity Related entities can be verified through methods such as digital signatures and hash verification. The identity legitimacy and behavioral compliance are verified to meet the following requirements: (1) 2) Transitivity: If entity trust , trust ,So You can also trust This transitivity allows trust relationships to extend into a wider network.

[0031] 3) Traceability: The system must fully record the process of establishing trust relationships to ensure that any trust decision is traceable. All of these can be verified through audit logs. Backtracking to the root of trust: (2) In this embodiment, raw data from terminal devices is collected via an edge gateway during the data acquisition phase. Device authentication and data integrity verification are combined to ensure the trustworthiness of the underlying devices. The specific implementation method is as follows: (1) Device authentication and data integrity verification: Device authentication determines whether a device is correctly connected by comparing its device ID.

[0032] Data integrity is verified by calculating a comprehensive integrity score, which is obtained by averaging the field integrity score and the data length score.

[0033] Field integrity score: (3) in, For the set of required fields, For the actual set of fields, This is the current timestamp. For device identity ID, The raw data collected by the device.

[0034] Data length score: (4) in, This represents the actual length of the decompressed data. (Bytes) is the base length.

[0035] The overall integrity score is: (5) You can set thresholds to determine whether the data integrity is up to standard, or you can use integrity scores to determine the reliability of the data collection part, thereby tracing back and checking whether there are any problems with the data collection part.

[0036] (2) Data anomaly alarm: Calculate the mean and variance of the data sequence within the time window, then calculate the Z-score for each data point. Use the Z-score method to determine if a data point is an outlier; for example, if the Z-score value meets the following criteria... When the device data changes are deemed to be in an abnormal state, an alarm is triggered.

[0037] In this embodiment, the data stream collected from the edge gateway is received in the scheduling and control stage, and the photovoltaic power generation and load are predicted. The prediction method can adopt existing methods, such as using LSTM neural network and random forest model to perform 24-hour rolling prediction of photovoltaic power generation and load power respectively.

[0038] Taking into account multiple factors such as load forecasting results, grid constraints, and economics, an optimal dispatching scheme is generated to realize the intelligent operation of the virtual power plant. The optimal dispatching scheme is mainly achieved by adjusting energy storage. The system prioritizes energy storage discharge during off-peak hours and energy storage charging during peak hours. Net power is calculated at flat electricity prices. The optimization objective is to minimize the interaction cost with the grid. The final energy storage output adjustment result and the actual exchange power with the grid are obtained by optimizing the objective function, thus generating the dispatching scheme.

[0039] The power balance constraint compliance and energy balance are calculated separately, and then the reliability of the scheduling scheme is calculated to ensure the reliability of the intermediate layer service. The specific implementation process is as follows: (1) The peak-valley electricity pricing strategy in this embodiment is formulated as follows: (6) in, This indicates the electricity price at different times. Peak hour electricity price, Off-peak electricity pricing This refers to the flat electricity price during off-peak hours. , These are the peak time period set and the valley time period set, respectively.

[0040] The system employs different optimization strategies based on different time periods. During off-peak hours, energy storage discharge is prioritized, while during peak hours, energy storage charging is prioritized. Net power is calculated at flat electricity prices. (7) if ,but: (8) (9) in, Net power, which is the surplus / deficit of photovoltaic power generation minus the load power consumption, is used to adjust the energy storage charging / discharging plan; For the rated capacity of the energy storage system, This represents the maximum charge and discharge power of the energy storage system. For maximum charging capacity, For charging efficiency, This represents the output power of the photovoltaic system at time t. This represents the power demanded by the load at time t. This represents the charging and discharging power of the energy storage system at time t, with a positive value indicating charging and a negative value indicating discharging. It is the state of charge of the energy storage system at time t. , This represents the state of charge of the energy storage system at the previous moment. The change in the state of charge of the energy storage system from time t-1 to time t; It is the maximum state of charge allowed by the energy storage system.

[0041] The above formula represents the calculation of the maximum rechargeable power when there is an energy surplus, taking the energy storage system's charging power limit, the current available net power, and the acceptable charging amount of the battery (considering charging efficiency). The minimum of these three limiting factors.

[0042] if ,but: (10) (11) in, This represents the maximum discharge power of the energy storage system. This refers to the discharge efficiency.

[0043] The above formula represents the calculation of the maximum discharge power when energy is insufficient, taking into account the charging power limit of the energy storage system, the current power gap required, and the discharge capacity that the battery can provide (considering discharge efficiency). The minimum of these three limiting factors.

[0044] (2) With minimizing the interaction cost with the power grid as the optimization objective, an objective function is established. By optimizing the objective function, the final output adjustment result of the energy storage and the actual exchange power with the power grid are obtained, so as to maximize the revenue of the virtual power plant. That is, the transaction volume is adjusted according to the change of electricity price to obtain the maximum profit.

[0045] The objective function is as follows: (12) Constraints: (13) (14) (15) (16) in, This represents the output power of the photovoltaic system at time t. This represents the exchange power of the power grid at time t, with a positive value indicating the purchase of electricity from the grid and a negative value indicating the sale of electricity. This represents the power demanded by the load at time t. This represents the charging and discharging power of the energy storage system at time t, with positive values ​​indicating charging and negative values ​​indicating discharging. The maximum permissible grid interconnection power is expressed in kW.

[0046] Equation (13) is the power balance constraint, Equation (14) is the energy storage state of charge constraint, Equation (15) is the energy storage charge and discharge constraint, and Equation (16) is the grid interaction constraint. These constraints are designed to ensure the lifespan of the energy storage system and the stability of the grid.

[0047] (3) Calculate the power balance constraint compliance and energy balance respectively, and then calculate the reliability of the scheduling scheme to ensure the reliability of the intermediate layer service; the reliability of the scheduling scheme can ensure that the scheduling plan meets the power balance and energy balance constraints and can be passed to the market transaction module for correct execution. For example, a reliability threshold can be set. When the reliability of the scheduling scheme reaches the reliability threshold, the scheme is considered reliable. Otherwise, an alarm is output. At this time, it can be checked whether the system is affected by network attacks, resulting in the incorrect reception of the scheduling plan.

[0048] Specifically, the power balance constraint compliance is determined based on the ratio of the number of constraint violations to the total number of checks; the energy balance is determined based on the output power of the photovoltaic system, the grid exchange power, the charging and discharging power of the energy storage system, and the load demand power at time t; the reliability of the dispatch scheme is the weighted sum of the power balance constraint compliance and the energy balance.

[0049] As a specific example, the calculation method for power balance constraint compliance is as follows: (17) The energy balance is calculated as follows: (18) The reliability of the scheduling scheme is calculated as follows: (19) in, This represents the degree of compliance with power balance constraints. It refers to the number of times the constraint was violated. This is the total number of checks. For energy balance, To assess overall credibility, and These are the weights assigned to the constraint compliance degree and the energy balance degree, respectively, and the sum of the two is 1.

[0050] In this embodiment, buy and sell orders are generated according to the scheduling scheme in the market transaction process, and a hash value is added to each transaction. A blockchain smart contract verification mechanism is adopted, and distributed ledger technology is used to realize the reliable execution of the top-level business.

[0051] (1) In the market transaction process, the order generation model simulates the actual operation process to generate buy and sell orders for the transaction. The matching model matches the buy and sell orders according to the matching logic. If the matching is successful, the transaction is successful. The revenue after the order generated by this scheduling plan is completed is obtained through the revenue calculation model.

[0052] The logic of the order generation model in this embodiment can be: when the power exchange of the power grid at time t... When, an electricity sales order is generated; when At that time, an electricity purchase order is generated.

[0053] The base electricity price set based on the peak-valley electricity pricing strategy is as follows: (20) in, Peak-hour electricity pricing taking market depth into account, To take into account off-peak electricity pricing based on market depth, Electricity prices for other time periods after taking market depth into account.

[0054] The above formula is used to generate prices during market transactions. It combines weighting parameters to generate the final transaction price, simulating real market transactions. For example, a base peak electricity price can be assumed. =1 yuan / kWh, current market price =1.2 yuan / kwh, and then the buyer and seller can calculate the optimal price according to formula (21) and (22).

[0055] The optimal price generation model that considers market depth information is as follows: (twenty one) (twenty two) in, It is an optimized offer from the seller after taking into account in-depth market information. It is an optimized offer made by the buyer after taking into account in-depth market information; The highest bid price for market depth. The lowest sell order price to incorporate market depth. , , , These are weighting coefficients, which can be used to adjust buy and sell order pricing strategies and can be set according to actual needs.

[0056] This embodiment first obtains the real-time market price based on formula (20), and then generates buy and sell order prices based on formulas (21) and (22). After that, it determines whether the two parties' quotations can be matched according to the order matching conditions, and the transaction volume is calculated according to formula (25).

[0057] The matching model is as follows: (twenty three) (twenty four) (25) The requirement for order matching is the purchase price. Greater than the sell order price 90% of the orders are matched, with a 5% price tolerance to allow more orders to be fulfilled. Transaction price The average of the buy and sell prices of successfully matched orders, and the quantity. The regulations specify the quantity of electricity purchased. Electricity purchase quantity with sell order The minimum value in.

[0058] The specific profit calculation model is as follows: (26) (27) (28) (29) in, For energy storage arbitrage profits, equation (27-29) represents the constraints that must be satisfied to maximize profits, where... For charging power, Equation (27) describes the change in the state of charge of the energy storage system between adjacent time nodes, while Equation (29) restricts the energy storage system from being charged and discharged simultaneously at the same time.

[0059] (30) (31) In the formula, Net income from power grid transactions Calculated for total revenue; Let t be the base electricity price. Let be the sell order price at time t. Let t be the purchase price at time t.

[0060] (2) Adopt a blockchain smart contract verification mechanism and use distributed ledger technology to realize the trusted execution of top-level business.

[0061] The blockchain model is the foundation for ensuring the trustworthiness of market transaction modules. Every order generated will have a corresponding hash value generated through the blockchain. Every order matched will also have a corresponding hash value, and the accepted scheduling plan will also have a corresponding hash value. The trustworthiness of transactions is ensured by comparing hash values.

[0062] A blockchain consists of a series of blocks, each containing several transactions and the hash value of the previous block, forming a chain structure.

[0063] (32) (33) (34) (35) Equation (32) defines the block. For block index, This is the collection of transactions contained in this block. For timestamps, The hash value of the previous block. The random number adjusted in the proof-of-work mechanism is determined by adjusting... The value of makes the hash value of the block meet certain difficulty conditions, thus ensuring the security and immutability of the blockchain.

[0064] The blockchain validity verification is set up to satisfy equation (34): the pre-hash of each block is correct; and equation (35): the proof-of-work of each block is valid.

[0065] (36) (37) (38) in, For the unconfirmed transaction pool, unconfirmed transactions will be... Add a new block Finding a legitimate one through proof of work Then the new block is added to the blockchain; This represents the last transaction in the transaction pool; This represents the index number of the last block in the current blockchain, used to determine the index number of the new block (new block index = last block index + 1); The current timestamp records the specific time when a new block is created, ensuring the temporal order of the blockchain and facilitating the tracking of block creation time; This represents the hash value of the last block.

[0066] In this embodiment, the credibility in a specified network attack environment is calculated through real-time trust assessment and adaptive weight adjustment to verify the ability of the trust chain to improve the credibility of the virtual power plant, thereby assisting the virtual power plant in multi-agent collaborative decision-making in a trusted environment for data collection, scheduling control, and market transactions.

[0067] This embodiment's real-time trust calculation framework includes three trust calculation dimensions and introduces a dynamic parameter update mechanism, constructing a complete trust assessment chain. This chain transforms the system from abstract power system nodes to specific industrial equipment and then to network attack targets, ultimately forming a complete transformation of the trust assessment subject. The final results show that introducing a trust chain into the virtual power plant significantly improves the system's credibility, allowing multiple subjects to make trustworthy decisions in a more reliable environment.

[0068] In this embodiment, subjective trust is determined based on direct trust value and recommended trust value; objective trust is determined based on attack success rate, security incident impact and vulnerability impact; subjective trust and objective trust are weighted and summed to obtain comprehensive trust; wherein, the weight coefficient is dynamically adjusted based on the number of interaction history; the comprehensive trust is adjusted through a trust evolution factor; the value of the trust evolution factor is determined based on the trust change trend.

[0069] Specifically, subjective trust is calculated as follows: (39) in, Subjective trust level, For direct trust weights, As a direct trust value, To recommend trust weight, Recommended trust value.

[0070] Objective trust is calculated as follows: (40) In the formula, To ensure objective trustworthiness, To increase the success rate of the attack, SuccessfulAttacks represents the number of attacks that successfully breached the defense, while TotalAttacks represents the total number of attacks executed. These two parameters are statistical results from actual system runtime. To assess the impact of security incidents, the severity of each incident is evaluated and mapped to a specific impact weight. Specifically, the system categorizes security incidents into four severity levels, each assigned a different weight: low risk (0.02), medium risk (0.05), high risk (0.10), and severe risk (0.15). For a set of security incidents, the overall impact is the average of all incident weights, with a cap of 0.3 to prevent excessive impact. This can be expressed mathematically as follows: ;in denoted as the weight of the i-th event, and n is the total number of security events. Security events refer to specific events that affect system security (such as unauthorized login, configuration errors, suspicious access, detection of malware, persistent attack attempts, data leakage, successful intrusion, etc.). In this system, they are divided into four levels according to severity: low, medium, high, and severe. The impact of vulnerabilities is calculated based on CVSS scores (directly calling the cvss library in Python). The system first calculates the average CVSS score for all vulnerabilities, then divides this average by 50 to scale it to a range suitable for trust level calculation, setting an upper limit of 0.2 to ensure the impact of vulnerabilities remains within a controllable range. This calculation process can be expressed by the following mathematical formula: ,in Let be the CVSS score of the i-th vulnerability, and n be the total number of vulnerabilities. The CVSS score is scaled by dividing by 50 to avoid a single factor excessively influencing the trust level.

[0071] The overall trust level is calculated as follows: (41) in , These are subjective trust weights and objective trust weights, respectively, and a dynamic weight adjustment mechanism is adopted. When the system lacks interaction history, the subjective trust weight is set to 0 and the objective trust weight is set to 1.0. When there is sufficient interaction data, the configured weights are used: for example, the subjective trust weight can be set to 0.6 and the objective trust weight to 0.4 for weighted average.

[0072] The final trust level is calculated by integrating trust data, and then adjusted using the trust evolution factor E. (42) (43) In this embodiment, the trust evolution factor E is based on the overall trust change trend. The trust score is calculated to be 1.0 when the system is stable, increasing by a maximum of 20% when trust increases and decreasing by a maximum of 30% when trust decreases, ensuring that the trust level remains within the range of [0,1] and reflects the true security status of the system. A comprehensive trust score and security report are then generated.

[0073] Application examples: To verify the effectiveness of the method in this embodiment, a simulation environment was built on the Python 3.12 platform, and a virtual power plant simulation system with a 24-hour operating cycle was designed. A complete trust chain covering three links—data acquisition, scheduling control, and market transactions—was constructed. Attacks were carried out on the IEEE 30-node network with the introduced trust chain and the standard IEEE 30-node network, and a mutual trust calculation analysis method based on mobile agents was used as an evaluation tool.

[0074] In the data acquisition phase, the system incorporates four core device types: smart meters, photovoltaic inverters, energy storage converters, and load controllers, generating a historical dataset covering a two-year time span, updated every 5 minutes. The edge layer constructs a complete edge computing architecture, including edge gateways supporting Modbus and MQTT protocols, edge computing modules for real-time data analysis, and an anomaly detection algorithm and data integrity verification model based on Z-score, ensuring data quality while storing encrypted data in the InfluxDB time-series database.

[0075] In the scheduling and control phase, the core scheduling mechanism uses asynchronous threads to retrieve the latest data from the InfluxDB database every 5 minutes, ensuring that the main scheduling process is not blocked. The system employs both LSTM neural networks and random forest models to perform 24-hour rolling forecasts of photovoltaic power generation and load power. Scheduling optimization aims to minimize grid interaction costs, implementing peak-valley pricing strategies and strictly adhering to operational constraints such as power balance constraints, energy storage SOC and charging / discharging power limits, and grid interaction power constraints. The system ultimately outputs hourly data on energy storage power, grid power, load power, and photovoltaic power, generating a basic scheduling plan every hour and updating it every 5 minutes.

[0076] In the trading platform phase, the system constructs a blockchain P2P trading environment based on the SHA-256 hash algorithm and proof-of-work consensus mechanism, employing a three-layer architecture: a blockchain interface layer, a market trading layer, and a core trading system layer. The platform establishes a time-of-use (TOU) electricity pricing system: peak hours (8-11 AM and 6-9 PM) are priced at 1.5 yuan / kWh; off-peak hours (other times) are priced at 0.8 yuan / kWh; and off-peak hours (0-6 AM and 10-12 AM) are priced at 0.3 yuan / kWh. The trading mechanism uses a heap data structure to prioritize orders, ensuring that orders with the best price are matched first, and triggering a mid-price matching algorithm when the buy price is not lower than 90% of the sell price. Every 5 minutes, the system writes the optimized scheduling plan to the blockchain, recording complete transaction hash values, timestamps, transaction status, and other information, achieving reliable verification and traceability throughout the entire transaction process.

[0077] Network Attack Phase: Three network attack scenarios were set up, as shown in Table 1. The Tiny scenario specifically targets the six generator nodes (nodes 1, 2, 5, 8, 11, and 13) in the system. These nodes are the core of the power system, controlling the power supply of the entire system. The attack complexity is set to a low level, mainly to obtain basic control commands, system status, and generation data. The Small scenario expands the attack scope to generator nodes plus some critical load nodes, increasing the attack complexity to a medium level. Target data types include generation data, protection settings, network topology, and load data. The Medium scenario is the most comprehensive attack strategy, targeting all important nodes (including all generator nodes and critical load nodes). The attack complexity reaches a high level, attempting to obtain all critical data of the system, including measurement data, status estimates, security policies, and backup files. The attack system integrates the NASim network attack simulation environment, employing a brute-force proxy for single-round intensive attacks, a multi-round intensive attack mode executing 200 rounds of continuous attacks, and an adaptive multi-round attack mode executing 150 rounds of intelligent attacks. A complete node-device mapping mechanism has been established, with generator nodes mapping to a combination of power generation-side equipment, including photovoltaic inverters, energy storage PCS, and smart meters, and load nodes mapping to important power consumption-side equipment, including loads and smart meters. This achieves a complete transformation from abstract power system nodes to specific industrial equipment, then to network attack targets, and finally to trust assessment subjects. Figure 3 This is a network topology diagram of the IEEE 30-node power system.

[0078] Table 1 Network Attack Scenario Settings

[0079] First, two years of historical data were initially generated for the prediction of photovoltaic power generation and load electricity consumption. The real-time generated data was written to and the historical data was updated every five minutes. Table 2 shows the real-time abnormal situation monitoring of photovoltaic equipment during operation.

[0080] Table 2 Equipment Operation Status and Anomaly Monitoring Table

[0081] As the sun gradually sets, the photovoltaic output decreases, and the energy storage begins to discharge. This screen displays an abnormal temperature change in the photovoltaic power generation equipment. Due to reduced sunlight, the equipment output decreases and the temperature drops; the change exceeds the specified threshold, triggering an alarm.

[0082] Then, the current time prediction module obtains the load and photovoltaic power predictions for the next 24 hours based on historical data, as well as the grid switching power, electricity price, and battery state of charge scheduling plan, such as... Figure 2 As shown, the photovoltaic and load power forecasts are within a reasonable range, and a reasonable scheduling plan can be generated based on the forecast results. Furthermore, the visualized scheduling plan generation provides a more intuitive view of whether the scheduling plan is feasible and whether the system is operating normally.

[0083] Similarly, the scheduling control module also has monitoring and evaluation functions. Table 3 shows some system monitoring data and scheduling scheme monitoring tables.

[0084] Table 3 System Data and Scheduling Scheme Monitoring Table

[0085] Subsequently, the market trading module receives the execution results from the scheduling module, generates buy and sell orders according to the scheduling plan, and adds a hash value to each transaction to ensure that transaction records are searchable and to guarantee transaction transparency and fairness. Table 4 shows an example of a blockchain transaction record: Table 4 Examples of Blockchain Transaction Records

[0086] Table 5 presents an economic benefit analysis of the scheduling plan generated based on the trusted verification subsystem for scheduling control. A comparison of the returns from energy storage arbitrage and direct trading with virtual power plants is also included. The table shows that the returns significantly increased after incorporating energy storage-optimized scheduling and trading, changing from a profit of 737.22 yuan to a profit of 1336.19 yuan.

[0087] Table 5 Economic Benefit Analysis

[0088] Finally, the system's credibility enhancement effect was tested under three attack scenarios. After a successful attack, the system assesses the number of compromised nodes, the leaked critical data, and the discovered security vulnerabilities, and calculates the cascading impact through network topology analysis.

[0089] In this embodiment, the initial values ​​of the direct trust value and the recommended trust value are first set, for example, both are set to 0.5; then they are updated based on the following formula. The recommended trust level has been updated to: (44) Direct trust level updated to: (45) Overall trust level updated to: (46) in, The recommendation trust update coefficient is used to control the system's sensitivity to third-party recommendation information. The direct trust update coefficient is used to adjust the system's learning speed from direct interactive experiences. The overall trust update coefficient is used to manage the dynamic evolution of the overall trust level. For time frames. Let this be the recommended trust value at time n+1. The recommended trust observation value at time n; Let be the direct trust value at time n+1. For the direct trust observation at time n; This is the overall trust value at time n+1.

[0090] In this embodiment, the calculation logic for trust level is as follows: Set initial values ​​for the recommendation trust score and direct trust score at time n, and the overall trust score at time n-1. The initial values ​​are all 0.5; the subjective trust level and objective trust level at time n are calculated using formulas (39) and (40) respectively; the calculation results are substituted into formula (41) to calculate the comprehensive trust level at time n; and the comprehensive trust level is adjusted using formula (42) to obtain the result. The recommendation trust level is updated using formula (44). The direct trust update is obtained by using formula (45). The overall trust level update is obtained by calculating using formula (45). This process of iterative updating continues until the end.

[0091] To avoid the subjectivity issues of traditional parameter allocation, and to evaluate the rationality of parameter selection and the robustness of results, this embodiment performs sensitivity analysis. In a simulation environment, this embodiment tests the coefficients... , , The impact of different combinations in the range [0.1, 0.9] on the trust score calculation results.

[0092] First, calculate the importance and interaction of the parameters: ; ; ; ; in, It is a first-order sensitivity index. It is the total effect index. It is an interaction effect. It's a trust score. Represents variance. Expressing expectations, Indicates except All parameters except those mentioned above; This represents the parameter currently being analyzed, which is the primary variable studied in the sensitivity analysis. Is it except Another parameter, Used to measure the strength of the interaction between two parameters, reflecting the parameters and parameters The degree of influence of both on the output variable Y.

[0093] By conducting sensitivity analysis and robustness testing on the three parameters, the subjective problem of parameter allocation can be avoided, making the credibility calculation results more representative and able to represent the credibility of the virtual power plant with this configuration in this scenario.

[0094] The performance evaluation function calculates a comprehensive score to select the optimal parameter combination and calculate the most representative credibility result. Finally, the parameter combination with the highest score is selected to calculate the system's credibility in this scenario. The specific performance evaluation function is as follows: ; in, It is a comprehensive score. It is a stability score. It is a robustness score. and The weights represent stability and robustness, respectively, with each baseline weight set to 0.5 and adjusted based on the sensitivity analysis results. The adjustment scheme is as follows: ; ; ; in, , It is the baseline weight. It is an interaction effect correction factor. This represents the strength of the interaction effect between parameters, calculated as the average of all parameter pair interactions. Specifically, when performing sensitivity analysis on the three update coefficients β(τ), λ(τ), and σ(τ), the interaction effects between all possible pairwise parameter pairs are calculated, and then the average of all interaction effects is taken. A comprehensive score is obtained by normalizing the sensitivity and adding an interaction factor correction.

[0095] The stability score is calculated using the coefficient of variation (CV) and parameter sensitivity: ; The coefficient of variation (CV) is a standardized statistical indicator used to measure the dispersion (volatility) of data. , It is the standard deviation of the data. It represents the average value of the data. For example, when calculating the trust score for 1000 generated sets of parameters, a set of 1000 comprehensive trust score values ​​is obtained. Calculating the mean and standard deviation of this set yields the coefficient of variation (CV). A high CV value means that the trust score is highly sensitive to parameter changes; even minor parameter adjustments can cause significant fluctuations in the trust score, leading to system instability. A low CV value means that the trust score is not sensitive to parameter changes; even if the parameters change within a certain range, the trust score remains within a stable range, indicating a very stable system.

[0096] Robustness scoring is based on the ratio of parameter changes to changes in confidence values: ; in, This refers to the change in overall trust level, specifically the difference in overall trust level calculated based on different combinations of parameters. This represents the change in the parameter.

[0097] Finally, malicious recommender detection: ; in, This represents the current recommendation trust value. Represents the direct trust value at the current moment; when When the threshold is exceeded, the recommendation is marked as malicious, and its recommendation trust weight is reduced to 0.

[0098] Through a comprehensive evaluation of system stability and robustness, a complete search space containing 1000 parameter combinations was constructed using Latin hypercube sampling for optimization comparison, in order to find the most representative parameter combinations and enhance the persuasiveness of the case study.

[0099] The final comprehensive trust calculation adopts a dynamic weight adjustment mechanism and is ultimately adjusted through a trust evolution factor to generate a comprehensive trust score and system security report.

[0100] Figure 4 A comparative analysis of the credibility calculation results after introducing the trust chain mechanism is presented. As shown in the figure, the system's credibility was significantly improved after introducing the trust chain mechanism in three different attack scenarios. In the tiny attack scenario, the credibility increased from 0.110 to 0.291, an increase of 164.2%; in the small attack scenario, the credibility increased significantly from 0.264 to 0.829, an increase of 213.9%; and in the medium attack scenario, the credibility increased from 0.372 to 0.847, an increase of 127.7%. Overall, the trust chain mechanism demonstrated good protection effects in all attack scenarios, with the most significant improvement in the small scenario. This indicates that the proposed trust chain mechanism can effectively enhance the system's credibility and security when facing network attacks of varying intensities.

[0101] Through the trusted verification architecture for virtual power plant aggregation operation in this embodiment, virtual power plants can not only achieve efficient monitoring and resource optimization of power grid operation, but also ensure the security, transparency, and immutability of information flow through blockchain technology, thereby solving the trust issues between virtual power plants and the power grid dispatching system, as well as in the electricity market. Ultimately, a trust network composed of multiple efficient, secure, and transparent virtual power plant trust chains is constructed, providing solid technical support for fair competition, intelligent dispatch, and reliable operation in the electricity market, ensuring efficient interaction and stable operation of the power system, and addressing problems such as... Figure 5 As shown.

[0102] Example 2 In one or more embodiments, a virtual power plant end-to-end trusted aggregation system based on a hierarchical trust chain is disclosed, comprising: The model building module is configured to: build differentiated hierarchical trust chain models for the data acquisition link, the dispatch control link, and the market transaction link based on the characteristic requirements and trust transmission rules of different business links of the virtual power plant; The data acquisition and trust verification module is configured to: collect raw data from terminal devices through the edge gateway during the data acquisition process, and combine device identity authentication and data integrity verification to ensure the trustworthiness of the underlying devices; The scheduling control trust verification module is configured to: receive data streams collected from the edge gateway in the scheduling control process, predict photovoltaic and load, and generate the optimal scheduling scheme; with the goal of minimizing the interaction cost with the power grid, calculate the power balance constraint compliance degree and energy balance degree respectively, and then calculate the trustworthiness of the scheduling scheme to ensure the trustworthiness of the intermediate layer service; The market transaction trust verification module is configured to generate buy and sell orders according to the scheduling scheme during the market transaction process, add a hash value to each transaction, and use a blockchain smart contract verification mechanism to achieve trustworthy execution of the top-level business. The collaborative decision-making credibility calculation module is configured to calculate the credibility of multi-agent collaborative decision-making through real-time trust assessment and adaptive weight adjustment.

[0103] In other embodiments, a terminal device is also disclosed, which includes a processor and a memory, the processor being used to implement instructions; the memory being used to store multiple instructions adapted to be loaded and executed by the processor to perform the above-described trusted aggregation method for the entire process of a virtual power plant based on a hierarchical trust chain.

[0104] In other embodiments, a computer-readable storage medium is also disclosed, which stores a plurality of instructions adapted to be loaded and executed by a processor of a terminal device, the above-described trusted aggregation method for the entire process of a virtual power plant based on a hierarchical trust chain.

[0105] The specific implementation process of the above modules or methods is the same as that in Example 1, and will not be described in detail again.

[0106] While the specific embodiments of the present invention have been described above in conjunction with the accompanying drawings, this is not intended to limit the scope of protection of the present invention. Those skilled in the art should understand that various modifications or variations that can be made by those skilled in the art without creative effort based on the technical solutions of the present invention are still within the scope of protection of the present invention.

Claims

1. A method for end-to-end trusted aggregation of virtual power plant processes based on a hierarchical trust chain, characterized in that, include: Based on the characteristic requirements and trust transmission patterns of different business segments of a virtual power plant, differentiated hierarchical trust chain models are constructed for the data acquisition segment, the dispatch control segment, and the market transaction segment, respectively. In the data acquisition phase, raw data from terminal devices is collected through edge gateways, and device identity authentication and data integrity verification are combined to ensure the trustworthiness of the underlying devices. The scheduling and control stage receives data streams collected from the edge gateway, forecasts photovoltaic power and load, and generates the optimal scheduling scheme. The optimization objective is to minimize the interaction cost with the power grid in order to maximize the revenue of the virtual power plant; the power balance constraint compliance degree and energy balance degree are calculated respectively, and then the reliability of the dispatch scheme is calculated to ensure the reliability of the intermediate layer service. In the market transaction process, buy and sell orders are generated according to the scheduling plan, and a hash value is added to each transaction. The blockchain smart contract verification mechanism is used to realize the reliable execution of the top-level business. The credibility calculation of multi-agent collaborative decision-making is achieved through real-time trust assessment and adaptive weight adjustment.

2. The method for end-to-end trusted aggregation of a virtual power plant based on a hierarchical trust chain as described in claim 1, characterized in that, Using TPM as the hardware root of trust in the central control area of ​​the virtual power plant, a tree structure is used to build the trust chain in the data acquisition stage; a star structure is used to build the trust chain in the scheduling and control stage, so that the central node of the star trust chain can establish a direct connection with each execution node; and a distributed P2P structure is used to build the trust chain in the market transaction stage.

3. The method for end-to-end trusted aggregation of a virtual power plant based on a hierarchical trust chain as described in claim 1, characterized in that, The trustworthiness of underlying devices is ensured by combining device identity authentication and data integrity verification, specifically as follows: Device authentication determines whether a device is correctly connected by comparing its device ID. Data integrity verification calculates field integrity score and data length score separately, and then calculates the average of the two to obtain the comprehensive integrity score; Calculate the mean and variance of the data sequence within the time window, then calculate the Z-score for each data point, and determine whether a data point is an outlier based on the Z-score.

4. The method for end-to-end trusted aggregation of a virtual power plant based on a hierarchical trust chain as described in claim 1, characterized in that, The optimization objective is to minimize the interaction cost with the power grid in order to maximize the revenue of the virtual power plant. Specifically: ； in, This represents the exchange power of the power grid at time t, with a positive value indicating the purchase of electricity from the grid and a negative value indicating the sale of electricity. This indicates the electricity price at different times of the day; the price will vary depending on the peak and off-peak hours.

5. The method for end-to-end trusted aggregation of a virtual power plant based on a hierarchical trust chain as described in claim 1, characterized in that, The power balance constraint compliance and energy balance are calculated separately, and then the reliability of the scheduling scheme is calculated to ensure the reliability of the intermediate layer service. Specifically: The compliance rate with power balance constraints is determined based on the ratio of the number of constraint violations to the total number of checks. The energy balance is determined based on the output power of the photovoltaic system, the grid switching power, the charging and discharging power of the energy storage system, and the load demand power at time t. The reliability of the scheduling scheme is the weighted sum of the compliance with power balance constraints and the reliability of energy balance.

6. The method for end-to-end trusted aggregation of a virtual power plant based on a hierarchical trust chain as described in claim 1, characterized in that, In the market transaction process, the order generation model simulates the actual operation process to generate buy and sell orders for trading. The matching model matches the buy and sell orders according to the matching logic. If the matching is successful, the transaction is successful. The revenue after the order generated by this scheduling plan is completed is obtained through the revenue calculation model.

7. The method for end-to-end trusted aggregation of a virtual power plant based on a hierarchical trust chain as described in claim 1, characterized in that, The credibility calculation for multi-agent collaborative decision-making is achieved through real-time trust assessment and adaptive weight adjustment, specifically as follows: Subjective trust level is determined based on direct trust value and recommended trust value; Determine the objective level of trust based on attack success rate, security incident impact, and vulnerability impact. The subjective trust level and the objective trust level are weighted and summed to obtain the comprehensive trust level; the weighting coefficients are dynamically adjusted based on the number of interaction histories. The overall trust level is adjusted by a trust evolution factor; the value of the trust evolution factor is determined based on the trend of trust changes.

8. A virtual power plant end-to-end trusted aggregation system based on a hierarchical trust chain, characterized in that, include: The model building module is configured to: build differentiated hierarchical trust chain models for the data acquisition link, the dispatch control link, and the market transaction link based on the characteristic requirements and trust transmission rules of different business links of the virtual power plant; The data acquisition and trust verification module is configured to: collect raw data from terminal devices through the edge gateway during the data acquisition process, and combine device identity authentication and data integrity verification to ensure the trustworthiness of the underlying devices; The scheduling control trust verification module is configured to: receive data streams collected from the edge gateway in the scheduling control process, predict photovoltaic and load, and generate the optimal scheduling scheme; with the goal of minimizing the interaction cost with the power grid, calculate the power balance constraint compliance degree and energy balance degree respectively, and then calculate the trustworthiness of the scheduling scheme to ensure the trustworthiness of the intermediate layer service; The market transaction trust verification module is configured to generate buy and sell orders according to the scheduling scheme during the market transaction process, add a hash value to each transaction, and use a blockchain smart contract verification mechanism to achieve trustworthy execution of the top-level business. The collaborative decision-making credibility calculation module is configured to calculate the credibility of multi-agent collaborative decision-making through real-time trust assessment and adaptive weight adjustment.

9. A terminal device comprising a processor and a memory, the processor for implementing instructions; the memory for storing multiple instructions, characterized in that, The instructions are adapted to be loaded by a processor and executed as described in any one of claims 1-7, which is a trusted aggregation method for the entire process of a virtual power plant based on a hierarchical trust chain.

10. A computer-readable storage medium storing a plurality of instructions, characterized in that, The instructions are adapted to be loaded and executed by the processor of the terminal device, and are based on the layered trust chain of the virtual power plant full-process trusted aggregation method as described in any one of claims 1-7.

Images